parallel programming & cluster computing mpi introduction
Post on 15-Jan-2016
50 Views
Preview:
DESCRIPTION
TRANSCRIPT
Parallel Programming Parallel Programming & Cluster Computing& Cluster Computing
MPI IntroductionMPI IntroductionHenry Neeman, University of OklahomaPaul Gray, University of Northern Iowa
SC08 Education Program’s Workshop on Parallel & Cluster computingOklahoma Supercomputing Symposium, Monday October 6 2008
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 2
What Is MPI?The Message-Passing Interface (MPI) is a standard for
expressing distributed parallelism via message passing.
MPI consists of a header file, a library of routines and a runtime environment.
When you compile a program that has MPI calls in it, your compiler links to a local implementation of MPI, and then you get parallelism; if the MPI library isn’t available, then the compile will fail.
MPI can be used in Fortran, C and C++.
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 3
MPI Calls
MPI calls in Fortran look like this: CALL MPI_Funcname(…, errcode)In C, MPI calls look like: errcode = MPI_Funcname(…);In C++, MPI calls look like: errcode = MPI::Funcname(…);
Notice that errcode is returned by the MPI routine MPI_Funcname, with a value of MPI_SUCCESS indicating that MPI_Funcname has worked correctly.
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 4
MPI is an APIMPI is actually just an Application Programming Interface
(API).
An API specifies what a call to each routine should look like, and how each routine should behave.
An API does not specify how each routine should be implemented, and sometimes is intentionally vague about certain aspects of a routine’s behavior.
Each platform has its own MPI implementation.
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 5
Example MPI Routines MPI_Init starts up the MPI runtime environment at the
beginning of a run. MPI_Finalize shuts down the MPI runtime environment at
the end of a run. MPI_Comm_size gets the number of processes in a run, Np
(typically called just after MPI_Init). MPI_Comm_rank gets the process ID that the current process
uses, which is between 0 and Np-1 inclusive (typically called just after MPI_Init).
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 6
More Example MPI Routines MPI_Send sends a message from the current process to some
other process (the destination). MPI_Recv receives a message on the current process from
some other process (the source). MPI_Bcast broadcasts a message from one process to all of
the others. MPI_Reduce performs a reduction (e.g., sum, maximum) of
a variable on all processes, sending the result to a single process.
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 7
MPI Program Structure (F90)PROGRAM my_mpi_program IMPLICIT NONE INCLUDE "mpif.h" [other includes] INTEGER :: my_rank, num_procs, mpi_error_code [other declarations] CALL MPI_Init(mpi_error_code) !! Start up MPI CALL MPI_Comm_Rank(my_rank, mpi_error_code) CALL MPI_Comm_size(num_procs, mpi_error_code) [actual work goes here] CALL MPI_Finalize(mpi_error_code) !! Shut down MPIEND PROGRAM my_mpi_program
Note that MPI uses the term “rank” to indicate process identifier.
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 8
MPI Program Structure (in C)#include <stdio.h>#include "mpi.h"
[other includes]
int main (int argc, char* argv[]){ /* main */ int my_rank, num_procs, mpi_error;
[other declarations]
mpi_error = MPI_Init(&argc, &argv); /* Start up MPI */
mpi_error = MPI_Comm_rank(MPI_COMM_WORLD, &my_rank);
mpi_error = MPI_Comm_size(MPI_COMM_WORLD, &num_procs);
[actual work goes here]
mpi_error = MPI_Finalize(); /* Shut down MPI */
} /* main */
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 9
MPI is SPMDMPI uses kind of parallelism known as Single
Program, Multiple Data (SPMD).This means that you have one MPI program – a single
executable – that is executed by all of the processes in an MPI run.
So, to differentiate the roles of various processes in the MPI run, you have to have if statements:
if (my_rank == server_rank) { …}
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 10
Example: Hello World
1. Start the MPI system.
2. Get the rank and number of processes.
3. If you’re not the server process:1. Create a “hello world” string.
2. Send it to the server process.
4. If you are the server process:1. For each of the client processes:
1. Receive its “hello world” string.
2. Print its “hello world” string.
5. Shut down the MPI system.
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 11
hello_world_mpi.c#include <stdio.h>#include <string.h>#include "mpi.h"
int main (int argc, char* argv[]){ /* main */ const int maximum_message_length = 100; const int server_rank = 0; char message[maximum_message_length+1]; MPI_Status status; /* Info about receive status */ int my_rank; /* This process ID */ int num_procs; /* Number of processes in run */ int source; /* Process ID to receive from */ int destination; /* Process ID to send to */ int tag = 0; /* Message ID */ int mpi_error; /* Error code for MPI calls */
[work goes here]
} /* main */
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 12
Hello World Startup/Shut Down[header file includes]
int main (int argc, char* argv[])
{ /* main */
[declarations]
mpi_error = MPI_Init(&argc, &argv);
mpi_error = MPI_Comm_rank(MPI_COMM_WORLD, &my_rank);
mpi_error = MPI_Comm_size(MPI_COMM_WORLD, &num_procs);
if (my_rank != server_rank) {
[work of each non-server (worker) process]
} /* if (my_rank != server_rank) */
else {
[work of server process]
} /* if (my_rank != server_rank)…else */
mpi_error = MPI_Finalize();
} /* main */
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 13
Hello World Client’s Work[header file includes]int main (int argc, char* argv[]){ /* main */ [declarations] [MPI startup (MPI_Init etc)] if (my_rank != server_rank) { sprintf(message, "Greetings from process #%d!“, my_rank); destination = server_rank; mpi_error = MPI_Send(message, strlen(message) + 1, MPI_CHAR, destination, tag, MPI_COMM_WORLD); } /* if (my_rank != server_rank) */ else { [work of server process] } /* if (my_rank != server_rank)…else */ mpi_error = MPI_Finalize();} /* main */
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 14
Hello World Server’s Work[header file includes]int main (int argc, char* argv[]){ /* main */ [declarations, MPI startup] if (my_rank != server_rank) { [work of each client process] } /* if (my_rank != server_rank) */ else { for (source = 0; source < num_procs; source++) { if (source != server_rank) { mpi_error = MPI_Recv(message, maximum_message_length + 1, MPI_CHAR, source, tag, MPI_COMM_WORLD, &status); fprintf(stderr, "%s\n", message); } /* if (source != server_rank) */ } /* for source */ } /* if (my_rank != server_rank)…else */ mpi_error = MPI_Finalize();} /* main */
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 15
How an MPI Run Works
Every process gets a copy of the executable: Single Program, Multiple Data (SPMD).
They all start executing it. Each looks at its own rank to determine which part of the
problem to work on. Each process works completely independently of the other
processes, except when communicating.
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 16
Compiling and Running% mpicc -o hello_world_mpi hello_world_mpi.c
% mpirun -np 1 hello_world_mpi
% mpirun -np 2 hello_world_mpi
Greetings from process #1!
% mpirun -np 3 hello_world_mpi
Greetings from process #1!
Greetings from process #2!
% mpirun -np 4 hello_world_mpi
Greetings from process #1!
Greetings from process #2!
Greetings from process #3!
Note: The compile command and the run command vary from platform to platform.
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 17
Why is Rank #0 the server? const int server_rank = 0;
By convention, the server process has rank (process ID) #0. Why?
A run must use at least one process but can use multiple processes.
Process ranks are 0 through Np-1, Np >1 .
Therefore, every MPI run has a process with rank #0.
Note: Every MPI run also has a process with rank Np-1, so you could use Np-1 as the server instead of 0 … but no one does.
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 18
Why “Rank?”
Why does MPI use the term rank to refer to process ID?In general, a process has an identifier that is assigned by the
operating system (e.g., Unix), and that is unrelated to MPI:% ps PID TTY TIME CMD 52170812 ttyq57 0:01 tcshAlso, each processor has an identifier, but an MPI run that
uses fewer than all processors will use an arbitrary subset.
The rank of an MPI process is neither of these.
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 19
Compiling and Running
Recall:% mpicc -o hello_world_mpi hello_world_mpi.c
% mpirun -np 1 hello_world_mpi
% mpirun -np 2 hello_world_mpiGreetings from process #1!
% mpirun -np 3 hello_world_mpiGreetings from process #1!Greetings from process #2!
% mpirun -np 4 hello_world_mpiGreetings from process #1!Greetings from process #2!Greetings from process #3!
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 20
Deterministic Operation?% mpirun -np 4 hello_world_mpiGreetings from process #1!Greetings from process #2!Greetings from process #3!
The order in which the greetings are printed is deterministic. Why?
for (source = 0; source < num_procs; source++) { if (source != server_rank) { mpi_error = MPI_Recv(message, maximum_message_length + 1, MPI_CHAR, source, tag, MPI_COMM_WORLD, &status); fprintf(stderr, "%s\n", message); } /* if (source != server_rank) */} /* for source */This loop ignores the receive order.
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 21
Message = Envelope+ContentsMPI_Send(message, strlen(message) + 1, MPI_CHAR, destination, tag, MPI_COMM_WORLD);When MPI sends a message, it doesn’t just send the contents;
it also sends an “envelope” describing the contents:Size (number of elements of data type)Data typeSource: rank of sending processDestination: rank of process to receiveTag (message ID)Communicator (e.g., MPI_COMM_WORLD)
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 22
MPI Data Types
C Fortran 90
char MPI_CHAR CHARACTER MPI_CHARACTER
int MPI_INT INTEGER MPI_INTEGER
float MPI_FLOAT REAL MPI_REAL
double MPI_DOUBLE DOUBLE PRECISION
MPI_DOUBLE_PRECISION
MPI supports several other data types, but most are variations of these, and probably these are all you’ll use.
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 23
Message Tags for (source = 0; source < num_procs; source++) { if (source != server_rank) { mpi_error = MPI_Recv(message, maximum_message_length + 1, MPI_CHAR, source, tag, MPI_COMM_WORLD, &status); fprintf(stderr, "%s\n", message); } /* if (source != server_rank) */ } /* for source */
The greetings are printed in deterministic order not because messages are sent and received in order, but because each has a tag (message identifier), and MPI_Recv asks for a specific message (by tag) from a specific source (by rank).
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 24
Parallelism is Nondeterministic for (source = 0; source < num_procs; source++) { if (source != server_rank) { mpi_error = MPI_Recv(message, maximum_message_length + 1, MPI_CHAR, MPI_ANY_SOURCE, tag, MPI_COMM_WORLD, &status); fprintf(stderr, "%s\n", message); } /* if (source != server_rank) */ } /* for source */
The greetings are printed in non-deterministic order.
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 25
Communicators
An MPI communicator is a collection of processes that can send messages to each other.
MPI_COMM_WORLD is the default communicator; it contains all of the processes. It’s probably the only one you’ll need.
Some libraries create special library-only communicators, which can simplify keeping track of message tags.
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 26
BroadcastingWhat happens if one process has data that everyone else needs
to know?For example, what if the server process needs to send an input
value to the others?MPI_Bcast(length, 1, MPI_INTEGER, source, MPI_COMM_WORLD);Note that MPI_Bcast doesn’t use a tag, and that the call is
the same for both the sender and all of the receivers.All processes have to call MPI_Bcast at the same time;
everyone waits until everyone is done.
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 27
Broadcast Example: SetupPROGRAM broadcast IMPLICIT NONE INCLUDE "mpif.h" INTEGER,PARAMETER :: server = 0 INTEGER,PARAMETER :: source = server INTEGER,DIMENSION(:),ALLOCATABLE :: array INTEGER :: length, memory_status INTEGER :: num_procs, my_rank, mpi_error_code
CALL MPI_Init(mpi_error_code) CALL MPI_Comm_rank(MPI_COMM_WORLD, my_rank, & & mpi_error_code) CALL MPI_Comm_size(MPI_COMM_WORLD, num_procs, & & mpi_error_code) [input] [broadcast] CALL MPI_Finalize(mpi_error_code)END PROGRAM broadcast
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 28
Broadcast Example: InputPROGRAM broadcast IMPLICIT NONE INCLUDE "mpif.h" INTEGER,PARAMETER :: server = 0 INTEGER,PARAMETER :: source = server INTEGER,DIMENSION(:),ALLOCATABLE :: array INTEGER :: length, memory_status INTEGER :: num_procs, my_rank, mpi_error_code
[MPI startup] IF (my_rank == server) THEN OPEN (UNIT=99,FILE="broadcast_in.txt") READ (99,*) length CLOSE (UNIT=99) ALLOCATE(array(length), STAT=memory_status) array(1:length) = 0 END IF !! (my_rank == server)...ELSE [broadcast] CALL MPI_Finalize(mpi_error_code)END PROGRAM broadcast
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 29
Broadcast Example: BroadcastPROGRAM broadcast IMPLICIT NONE INCLUDE "mpif.h" INTEGER,PARAMETER :: server = 0 INTEGER,PARAMETER :: source = server [other declarations]
[MPI startup and input] IF (num_procs > 1) THEN CALL MPI_Bcast(length, 1, MPI_INTEGER, source, & & MPI_COMM_WORLD, mpi_error_code) IF (my_rank /= server) THEN ALLOCATE(array(length), STAT=memory_status) END IF !! (my_rank /= server) CALL MPI_Bcast(array, length, MPI_INTEGER, source, & MPI_COMM_WORLD, mpi_error_code) WRITE (0,*) my_rank, ": broadcast length = ", length END IF !! (num_procs > 1) CALL MPI_Finalize(mpi_error_code)END PROGRAM broadcast
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 30
Broadcast Compile & Run% mpif90 -o broadcast broadcast.f90
% mpirun -np 4 broadcast
0 : broadcast length = 16777216
1 : broadcast length = 16777216
2 : broadcast length = 16777216
3 : broadcast length = 16777216
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 31
ReductionsA reduction converts an array to a scalar: for example,
sum, product, minimum value, maximum value, Boolean AND, Boolean OR, etc.
Reductions are so common, and so important, that MPI has two routines to handle them:
MPI_Reduce: sends result to a single specified process
MPI_Allreduce: sends result to all processes (and therefore takes longer)
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 32
Reduction Example
PROGRAM reduce IMPLICIT NONE INCLUDE "mpif.h" INTEGER,PARAMETER :: server = 0 INTEGER :: value, value_sum INTEGER :: num_procs, my_rank, mpi_error_code
CALL MPI_Init(mpi_error_code) CALL MPI_Comm_rank(MPI_COMM_WORLD, my_rank, mpi_error_code) CALL MPI_Comm_size(MPI_COMM_WORLD, num_procs, mpi_error_code) value_sum = 0 value = my_rank * num_procs CALL MPI_Reduce(value, value_sum, 1, MPI_INT, MPI_SUM, & & server, MPI_COMM_WORLD, mpi_error_code) WRITE (0,*) my_rank, ": reduce value_sum = ", value_sum CALL MPI_Allreduce(value, value_sum, 1, MPI_INT, MPI_SUM, & & MPI_COMM_WORLD, mpi_error_code) WRITE (0,*) my_rank, ": allreduce value_sum = ", value_sum CALL MPI_Finalize(mpi_error_code)END PROGRAM reduce
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 33
Compiling and Running% mpif90 -o reduce reduce.f90
% mpirun -np 4 reduce
3 : reduce value_sum = 0
1 : reduce value_sum = 0
2 : reduce value_sum = 0
0 : reduce value_sum = 24
0 : allreduce value_sum = 24
1 : allreduce value_sum = 24
2 : allreduce value_sum = 24
3 : allreduce value_sum = 24
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 34
Why Two Reduction Routines?
MPI has two reduction routines because of the high cost of each communication.
If only one process needs the result, then it doesn’t make sense to pay the cost of sending the result to all processes.
But if all processes need the result, then it may be cheaper to reduce to all processes than to reduce to a single process and then broadcast to all.
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 35
Non-blocking Communication
MPI allows a process to start a send, then go on and do work while the message is in transit.
This is called non-blocking or immediate communication.
Here, “immediate” refers to the fact that the call to the MPI routine returns immediately rather than waiting for the communication to complete.
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 36
Immediate Sendmpi_error_code =
MPI_Isend(array, size, MPI_FLOAT,
destination, tag, communicator, request);
Likewise:mpi_error_code =
MPI_Irecv(array, size, MPI_FLOAT,
source, tag, communicator, request);
This call starts the send/receive, but the send/receive won’t be complete until:
MPI_Wait(request, status);
What’s the advantage of this?
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 37
Communication Hiding
In between the call to MPI_Isend/Irecv and the call to MPI_Wait, both processes can do work!
If that work takes at least as much time as the communication, then the cost of the communication is effectively zero, since the communication won’t affect how much work gets done.
This is called communication hiding.
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 38
Rule of Thumb for Hiding
When you want to hide communication: as soon as you calculate the data, send it; don’t receive it until you need it.
That way, the communication has the maximal amount of time to happen in background (behind the scenes).
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 39
To Learn More
http://www.oscer.ou.edu/
Thanks for your attention!
Questions?
SC08 Parallel & Cluster Computing: MPI IntroductionOklahoma Supercomputing Symposium, October 6 2008 41
References
[1] P.S. Pacheco, Parallel Programming with MPI, Morgan Kaufmann Publishers, 1997.[2] W. Gropp, E. Lusk and A. Skjellum, Using MPI: Portable Parallel Programming with the Message-Passing Interface, 2nd ed. MIT Press, 1999.
top related